Artificial Intelligence in Healthcare Keynote Speech

Artificial Intelligence in Healthcare Keynote Speech

yesterday I was in Kansas City speaking for NAIC which was a group of all the Insurance Commissioners from across the country about six hundred people a really fascinating con conference and I do this is pretty much all I do I do 40 or 50 events every year in all sorts of different parts of the world in all sorts of different industries so I'm not an insurance specific person I'm in many different industries in many different parts of the world and one of the most fascinating things from I love what I do I love it and learning a little bit about all these different industries it's it's unbelievable that kind of the insights you can gain from that and I'm always amazed of course at how different all these different industries are in the different circumstances they have but at the same time what's maybe more surprising is how much they're really just the same and everyone's kind of facing the same challenges and a big part of that today is technology so I focus on technology trends that's why I've been following that I've been doing this full-time for 12 years and and this is ubiquitous across every industry in fact you know I'm always going to meet these different people in different conferences and so on and everyone always says we're way behind right it's not true everyone's kind of in the same place except for maybe Google and Apple and Facebook they're on the cutting edge but everyone else is fighting like crazy to try and keep up with the times and it's accelerating right so I see that message everywhere and I always tell people we have to stay aggressive we have to stay proactive but don't feel like you're behind I think in fact this particular audience is ahead you know in a whole bunch of different ways again I'm not in the group but I've heard a lot about it kind of from the outside but anyway I don't want to bore you with a whole bunch of charts today but I do want to show you a couple and the first one we've heard about this sort of thing for a long time but I want to just really highlight the scale of what we're talking about and the topic here is the cost of storing one terabyte of data right simple enough so in the year 2000 the cost of storing one terabyte of data was approximately $17,000 so does anyone want to hazard a guess on what we're expected to pay next year is your point one cent okay you're gonna make my job art the right answer is three bucks so and then this is crazy it's not just happening in data storage it's happening in data bandwidth data processing and data storage which are the three pillars of technology that's why we're seeing this whole explosion is that the cost function is collapsing it's getting cheaper and cheaper right and so in the the change is affecting every industry and it's affecting every different product category and service category it's more and more your competitors in the future are all going to be technology companies I mean that's what it's becoming these technology companies are there they see what's happening because they own the platform right like you think about Amazon Web Services right they've got all these people running their businesses off of their servers they can see which business models are doing well with which ones aren't so it's very easy for them to cherry-pick the industries that are profitable doing well growing well suited to their expertise and knock them off one after another right so it's really an incredible thing and it's an it's an exponential environment and as human beings we are hardwired to think in linear terms so I want to talk about that in just a few minutes the exponential nature of this is incredible but the collapsing cost structure in technology that's of course the genesis behind the big data movement if it's cheaper to get the data and store the data and process the data then of course everyone's going to just try and get more and more of it now that's turned to a form of currency it's all about data this that's the beginning of the Internet of Things the IOT some people refer to it as M to M machine to machine estimates now that we're gonna actually beat the previous estimate of 50 billion devices connected by 2020 where we're butting up against that now so we're gonna surpass that and so there's all these sensors being developed for every possible application right there's all sorts of there's companies specialising in all this stuff and one of the most interesting areas to look at actually is the whole area of facilities management so even a room like this probably has a hundred sensors measuring all sorts of different things one of the best examples of facilities management today is the Hudson Yards facility going up in New York city anyone here from New York just said of curiosity what a great city the god architecture going up in New York right now it's totally redefining the skyline really interesting to see in Hudson Yards as a part of it 19 towers some of them are upper already it's going to be the most censored residential environment in the world when it finally opens but of course that title won't be held long because there's a million other projects in the world they're gonna be even more by the time they open right but with these added sensors right you can start to detect anomalies you can start to mitigate risk right and then if you've got the the sensors in the wearable tech space and increasingly some of that technology is starting to detect whether or not the people are smokers for example or if they're eating their vegetables right and a whole host of other medical metrics so this is absolutely gonna be fundamental in the insurance space I mean if this data is gonna have a dramatic impact already has right so you have startups coming into the space that are providing risk mitigating services right there's a whole host of that and by the way we're just getting started right because the deployment of 5g is gonna change the whole thing the bandwidth increases the the power required so what we're talking we're talking about sensors little plastic sensors that can in some cases they can get their power from a Wi-Fi network so they don't to be connected to anything they could use be stuck in a wall or a machine or whatever and they have enough power to measure whatever it is they're measuring and send the data these things are like a buck each even cheaper so we're gonna see an explosion of sensors like we've got 50 billion now we are just getting started there's gonna be sensors on everything going forward right and so then that they emphasis transitioned over into the dashboards and data visualization how can you interpret all this data there's incredible things going on I'm sure many of you have dashboards like this they are you looking at and so it allows people to start to understand what's going on right lowering the risk right how do we lower the risk and therefore make insurance cheaper across the board because there's fewer catastrophic events right and then we've seen a big revolution in the retail space of course retail is dying sadly late Amazon is completely changing the game and people talk about the Amazon effect which is an amazing thing right because we're all b2c and b2b people are starting to expect that higher levels of service that they've grown accustomed to with Amazon right and so you bring it into the FinTech space and we have these new players coming on board that are redefining the products you know bank accounts and investment accounts so we've had four generations but they're doing it in a new way right engaging with their customer and a cleaner portal right better optimized more intuitive portal and then you go into the insure tech space and you've got companies like hippo right redefining homeowners insurance engaging with the customers differently than we used to right in these other companies I mean this lemonade accelerating how fast we pay claims I mean it's unbelievable the amount of innovation that's happening and then even the qualification process is accelerating with companies these are all just startups and they were there this is not an exhaustive list there's so many of them out there but life insurance you answered 12 questions let AI do the rest and you just qualified right and then there's whole new categories of insurance as well which is the on-demand space and single item insurance all sorts of creative products where you can get insurance for literally just one specific thing dengue fever insurance for 10 bucks unbelievable right when in the peer-to-peer space insurance is coming online and finally just to kind of put a cap on the whole thing we've got a flood of innovation that's coming in from China where you know this is WeChat pay here and you've got Ally pay as well if China has 50 times more mobile payments per year than we have here in the u.s. 50 times not double not triple not tenfold 20 fold 50 times right so they've got way more data that's a huge advantage right and then you've got new players over there that are redefining the insurance pace online only so it truly is an exponential space right exponential environment that's what we're a part of that's what we get to live through right you guys I'm a left brain data nerd so I don't know how many of you are nerds or geeks but I'm one of them and these are exciting times like there is it can be intimidating but it can be exciting at the same time so I want to talk about that the exponential thing because you know I do this myself like I mean I literally spend all day every day trying to stay in front of these trends and understand what's going on and I always have to coach like know what if it was ten times as powerful as it is today like how would that change or better yet think about the the the the most expensive technology that you're implementing in your firms right now right what if that technology was one one-hundredth of the cost that you're paying how many of your competitors would use it then raise really and it's not a natural thing it's counterintuitive because the numbers get crazy too quick it's hard for your brain to to follow along so I have another example and again this is an example you've all heard before but but I want to really point out this scale of what's happening so of course I'm talking about the Human Genome Project which started in 1990 1990 I graduated high school in 1989 so I'm dating myself I'm 48 years old anyway they started it in 1997 years later they were one percent complete and all the pundits came out of the woodwork and they're like oh my gosh we're gonna need 700 years right if it took seven years together but of course that's not what happened and there's some people who are really good at thinking exponentially by the way Ray Kurzweil is is a very good example I'm sure some of you follow rayker as well for those of you who don't know he's the the guy who wrote the singularity is near and he's works for Google today eccentric guy but very interesting and extraordinarily accurate anyway so he said check this out he said if we're at 1% we're almost done that's fascinating if we're at 1% we're only if we're at 1% we're halfway there because if you're scaling at a hundred percent per year okay so it's doubling every year if you're scaling 100 percent per year how many years is it take to get from 1% to a hundred six and a half it's only six and a half years right one two two four eight sixteen thirty-two that's the fifth year sixty-four seventh year you're over the top like how's that possible I mean seriously when you think about it it's so crazy it doesn't seem possible and that's exactly how it played out in fact it we actually finished it slightly ahead of that schedule which means that it was scaling it a little bit more than a hundred percent each year now I'm not saying everything's scaling at that rate right Moore's law has slowed down dramatically solar power just for relative comparison accounts for about 0.8 percent of global energy production today but and it's scaling at about 25 percent per year meanwhile energy consumption is going up at about 1.8 percent per year anyway I'm being ad a math nerd but if you draw the lines out if the trends continue and we don't know if they will but if the trends continue then solar could feasibly satisfy global energy consumption by 2040 one will see we got to walk the path that's pretty incredible but anyway that's what's happening is these technologies are propagating at an accelerating pace so this is a graph of a whole bunch of different technologies that have come on board over the last hundred and twenty years and it shows when they you know from the Genesis when they were first deployed how long did it take to get up to most people count eighty percent as kind of broad market adoption so how long does it take to get to eighty percent and you can see that as we get to the present day they're just getting steeper and steeper these new technologies are coming really really quick so there's a lot of I mean of course this is this is affecting the length of the average tenure on the S&P 500 for example used to be like 65 years today it's 14 there was a study that came out of Washington University that said that 40 percent of sp500 companies are gonna be gone by 2026 is that possible I mean it's crazy when you think about and we don't know we have to I mean you kind of you can make predictions but it's you know in a lark to a large extent predicting the future as a fool's game but it's possible that we could get there and here's the problem and then this is I mean I see this face to face like every day the executives in all these companies in all these industries and they all say that they're behind and they're terrified right they're terrified that this is coming too quick and they're not going to be able to catch up and they're thinking inevitably behind every decision they're thinking who's gonna eat my lunch right who's gonna come and eat my life who's gonna steal my revenue and I think I mean this is like my whole like I mean I almost think it's my pertinence to make my mission it is exactly the wrong approach right what you and I know it's a flawed metaphor okay but what we should be saying is who else is lunch can we eat right because if you seriously if you think about it if if 40% of the SP 500 companies are gone and again I'm not going to debate whether or not that's an accurate prediction but let's just pretend that it plays out that way if for it's not like the economy is gonna shrink by 40 percent the economy is gonna be there so if these big companies falter and some of them who's gonna take their place who's gonna take their place that's what's happening all those startups I just showed a second to go with hippo and lemonade and all these people right these people are nipping in my gosh in this industry in particular because there's so much money involved right the financial incentive is huge so the and there's tons of investment coming in from everywhere I was it lived in the Bay Area for 18 years and the VC folks down on Sand Hill Road and private equity my gosh they're pouring money into this because they know if they hit one it's going to be a home run there's so much money involved right so all these companies are nipping in it the revenue that you've got right that's what's happening so we can play that game to the bigger companies can play that game too but we have to stay on offense right we have to stay on all we have to look for new revenue opportunities we're gonna talk about that later there's so much stuff my gosh change creates opportunity when other people are floundering and failing that's an opportunity to expend horizontally across the the industry chain vertically through the supply chain all sorts of stuff you can do so what I invite you to do this is a short session just take a step back I know you all have your I know you're all senior managers from what I understand c-suite and insurance companies but just take a step back and let me just share some of this stuff that I study this stuff every day I got so much I could keep you here for four days I have so much stuff but don't worry I won't I got a timer it's right here telling me how much time I got left so we're good but I want to start off with it with it with a quick story just out of curiosity how many of you recognize either of these two guys nice well we got one guy love it okay so for those of you who don't know the we'll start with the guy on the left his name is Brian Acton and in 2009 Brian Acton was looking for a job in the Bay Area and he applied to both Facebook and Twitter and they both turned him down and he's a great if you follow I've been following this guy for a long time he's actually had like a he's a nice guy he's a gracious guy and he wrote a tweet about it at the time which went like this he said Facebook turned me down a great opportunity to connect with some fantastic people looking forward to life's next adventure and what might that be don't say it he got together with a friend of his and started whatsapp which they grew right over the course of five years and sold to none other than Facebook for a cool 19 billion dollars so I'm bitter I admit it I'm 48 he's a lot younger than I am anyway so the pundits came out in full force again this is what the pundits do and there's a lot of different ways this was a record-breaking transaction I mean it was a big deal when this thing long I mean it was huge anyway so they came out there on the evening news and there's a lot of different ways you can look at this one way is that at the time of the sale whatsapp had 55 employees so by that measure by that measure Facebook paid 345 million dollars per employee incredible another way that is at the time of the sale what's up had an average of 450 million active monthly users so by that measure Facebook paid $42 per user either way it's a lot of money maybe the most interesting way to look at this transaction though is that right we're talking about five years like five years ago is 2014 2014 feels like yesterday to me I mean it just does you know it's crazy and so the 55 employees managed to engage over 450 million active monthly users over five years how many people are in this room like 250 maybe I don't know 250 555 like little group like that builds something in five years that touches the lives of a half a billion people and this speaks to the leverage that's in the system today and listen anytime you hear a speaker wherever conferences he or she always has a few like really key messages that they're hoping that you're gonna walk away with this is one of mine okay there's more and more leverage in the system all the time every year is more leverage its financial leverage different types of leverage but the biggest example in the room is technology technology is a form of leverage is in fact they one of the gentlemen on the panel just said that a second ago it's a lever right technologies of form so the stories that we hear of like the division between rich and poor and all that number one those stories are true it's it's it's widening quickly right now but a lot of it is because of this not all but a lot of it is because of this leverage is because some people and some companies are leveraging the technology and for the most part they're doing really well not all but many right and on the other side there's a ton of people and a ton of companies that are not leveraging the technology and in fact they're basically being leveraged by the technology right in other words other people are using technology against them right that's what these startups are doing and in those cases it's harder and harder every year to have the same performance that you had last year right you're swimming upstream in other words we have to run towards the technology we can't run away from it right data analytics is a form of leverage you can leverage data analytics we have to find as many uses as possible to leverage data analytics right 5g is a form of leverage we can leverage 5g right we can leverage robotics robotics my gosh is stuff showing up all over the place you see them in hotels you see them in shopping malls security robots and information robots and airports in fact those are leading indicators right there you can model the future because you've got these big institutions like stadiums and hospitals and shopping malls and whatever hotels airports they've got a larger financial incentive because they're processing higher numbers of people so they deploy the technology first or they test it first and we see if it works but we all can go to those sorts of facilities and see what technology they're using and have that as a proxy of what's likely to be used in smaller facilities smaller businesses and maybe eventually in people's homes down the road raise that so it's right that's where I always try to do is I try to find models how can we see what could be coming down the road right this is all part of the robotic process automation I'm sure you guys are all doing stuff in RPA it's crazy what what's happening there people think it's the future but it's already happened one of the best examples look at this this is the UBS trading floor 2005 2016 look at this the place used to be packed with people they're gone what happened to those jobs its algorithmic trading it's algorithmic trading platform it's all been automated right you don't need the people anymore right so these are examples of leverage but there's really two in particular that and again it's a short session so is it's it's two in particular I want to kind of dig a little bit deeper in and the first one is machine learning artificial intelligence machine learning right those two go hand in hand and the second is blockchain and you guys are in the cutting edge of blockchain you're doing some really cool stuff which i think is awesome some of that was told about it Andrew right yeah he was telling me just just an hour ago really fascinating but let's start with artificial intelligence so and this is where everyone always starts what constitutes artificial intelligence like most of it's just good analytics this is a good algorithm like what actually makes it artificial intelligence and there's generally two I mean people debate these things but there's generally two things that people look at and the first is that it should involve some level of human human level comprehension there something that appears human level and secondly more importantly it has to have the ability to learn that's the big one right does the platform have the ability to learn from the data that it's accumulating and other data that's added to the mix so what was the early winner of artificial intelligence well the big one was the recommendation engines and this is kind of old news at this point they just started you know eight years ago but you know you go to Amazon and it says you know you might be interested in these other things or if you're watching Netflix it'll suggest movies and the same thing on Pandora and so on but this is this is starting to come right into I mean it's not just insurance but just all the different industries cuz this is all driving getting people the right product at the right time it's just predictive analytics right understanding the customer journey and that's there's been a lot of like case histories I do a ton of stuff for credit unions I don't know how credit unions have so much money they have more I've used to think insurance had more conferences than any other industry I think credit unions have you beat they have they have conferences like I'm going to the Bahamas in July for a credit union conference unbelievable anyway they have case histories like am i complaining no but that stuff costs money right I mean it's incredible you should see the hotel anyway I love what I do did I mention that so there was a thing son-of-a-gun I forget anyway there was a credit union that all they did I mean listen to this simplicity is deceptive all they did was look at the you know the services that they offer products and services that they offer and they looked at him on a branch by branch basis assuming that different neighborhoods have different sales profiles is different demographics whatever this is natural that they would be buying different things so all they did was make the pops4 than signage in their credit unions match the product mix of what was actually selling in those credit unions and this sales increased 12% just by matching the signage how insane is that right so and you can get way more sophisticated than that but that's what's happening is understanding the customer journey have these people interacted with us in the past right where's that a good interaction a bad interaction did they buy something what are they likely how old are they what stage of life are they at and give them the right product to the right place you can optimize stuff right you can lower risk predictive analytics is being used to lower risk we mentioned that before and then here's the conversion funnels on you know you you you optimize your conversion funnel on your website you're lowering the cost of acquisition and in this industry if you lower the cost of acquisition that's everything right you own the whole space if you can lower the cost of acquisition because it's so darn high right and then predictive analytics is being used to hire people and what attributes are likely to lead to success in an insurance space right so this is happening all over the place it's all predictive analytics right that's a huge artificial intelligence piece but it's not the only one so you also have object and image recognition including facial recognition right the the they're using that now and the TSA or the you know the security deals in fact when I was just in Amsterdam and I went to board my flight and if you just walk up to it you know these Northern Europeans I mean bless their heart they're incredible people but they're like robots like they don't talk like people they're just like you go here and go you know but you go up and it's got the gate and it's got a camera it looks at your face and it knows if you're the right person the gates opened it's unbelievable right but object recognition is also being used in advanced manufacturing to for noticing defects and things and of course you get closer to home for this industry it's being used you know analyzing x-rays and all sorts of scans and the audience incredible with what artificial intelligence is able to learn it's getting better and better better than doctors in many cases they're looking at scans and predicting seizures and again in that on all these areas are getting better and better as time goes on right it's even being used believe it or not to identify victims of domestic abuse because a lot of times people come into an emergency room and they got an injury and they say it happened because you know if this happened it turns out it's not true and the humans can't always detect it but the machines can because they've got all the data from all the other people who have done similar things all across the country and across the world really fascinating right unbelievable and all sorts of medical imaging applications it's all object and image recognition right we can use it to lower risk it's even being used when people are submitting claims where the machines can look at the photograph and understand basically what it's looking at at least to be able to categorize it right it's not like it's a finished job but instead of requiring 20 man-hours to process their person hours I should say to process a hundred claims I don't know what those numbers are okay but maybe it only takes fifty person hours right and then maybe in another ten years it'll only take twenty right it just gets slowly better and better but where's the big one for object and image recognition right we've talked about a few but the huge one is in autonomous driving this is coming so much faster than people think it's crazy this is how quickly this is coming so and that's going to have a direct effect of course on car insurance and injuries as a result of track so I want to go into this I know it's not directly related but it's it's kind of indirectly related to your industry so fascinating so what's the what's the most complex environment for an autonomous vehicle right most obstacles most unforeseen situations well that's in city driving right that's where you're going to get the most number of obstacles right so who's making a play in that space well of course you've got Google's way mo and they've got about ten million miles of data driving data from their cars you've got about six thousand so they got roughly ten ten million miles of data to work with but then you got Tesla they have a billion miles of data accumulated because they got a hundred thousand cars on the road full of sensors right date is the gate guys everyone's looking for data right why did Facebook pain EIN teen billion for whatsapp they wanted the data and all the kids were leaving Facebook because their parents were on it you wanted to spend time on Instagram and whatsapp so what at Facebook do they bought Instagram and whatsapp now they've got the data right that data is worth everything what's the number one thing machine learning needs data right and so you've got these you've got these electric cars as crazies boobers making a big play and then if you go to China China is doing incredible things guys amazing was I speak over there it's incredible what they're doing unbelievable so this is the uber equivalent in China it's called didi they've got a huge play where they can do anything now because there's very little regulation that's the thing like you go to Europe they got gdpr right with the right to be forgotten and data privacy and everything and I respect that but then you go to the opposite side of the of the spectrum you got China there's no regulations they got a surveillance state they got more data than anybody so guess what they're machine learning is learning faster yeah I'm not saying we should go to that model but we should all be aware of the dichotomy between the two where the GDP are is a respectable effort but it's hindering the innovation when you compare it to China okay anyway how do these things learn and this is where it gets super fascinating so Tesla introduced their autopilot function in in 2015 that's when that when that started and and the hardware was already on the cars so there was like 50,000 cars or what I already had the hardware when they made the announcement so there was a guy I lived in the Bay Area for 18 years I moved south of there now but anyway he lives in the Santa Cruz Hills which is a beautiful area and on the way to his house there's like a really tight hairpin corner this isn't it I just bought an image okay but you get the idea so that first night they made the announcement and he's driving home from work approaching this tight corner and he took his hands off the wheel he was like all right let's see what you got not sure I would have done it but anyway so the car didn't crash like it you know did it but veered outside of the lane and of course the brakes came on and warning lights and the whole thing so it's basically I failed and then the next day he did it again and again and again and again two weeks later the car knew how to take the corner but it wasn't just his Tesla that knew how to take the corner it was every Tesla guys this is it right here seriously they call it fleet learning this is why it's so crazy okay you go to you go to Sand Hill Road on in the bay area of all the VC this is all they're talking about Network effect fleet learning right because if you can take the individual experience of individual people on the network and upload that data to the cloud that's the genius of the cloud where the true intelligence is up here then you can fix the problem and deploy network wide yeah we're quite as dude I'm a nerd seriously like I get shivers when I think about this stuff because anyway but it's true and so the this the in other words the improvements are accelerating at an accelerating rate I think it's going really quick people don't realize how quick it's going it's moving really quick so city drivings got the most number of obstacles right what's got fewer obstacles than city driving well you can go in the freeways that has less obstacles right so just about two years ago now they'd Auto did the POC proof of concept with Auto was purchased by uber so they ran this truck 125 miles down the freeway in Colorado there was a guy in the cab but he wasn't driving payload was fifty thousand cans of Bud your Budweiser got there on time and there's all sorts of you know 47% of trucking fees are the drivers salary 47% that means the first one to deploy has a two to one price advantage it's gonna go quick once the first person deploys it's gonna go fast right Tesla's got a model this is the Mercedes concept and Volvo so check this out if you don't need a driver you don't need a cab so they came out with this look at that I mean this is we this is stuff is it's coming I bet and look if I'm wrong you can call me out I'm cool I bet that at least one state in the Union in our country will legalize hub to hub long-haul autonomous trucking by the end of this year and if not by the end of this year next year for sure a USPS has got a POC running right now there's some of you are nodding you've heard that story they're doing it right now they're doing five trips to forget where they're going does anyone know anyway they're doing five round-trip trips there's going to be a safety driver in the cab but the person doesn't have to be driving pretty crazy it's going to come really fast know what's got fewer obstacles than freeways agricultural land right look at this thing this thing is badass like I would love to buy one of these myself I mean I'm not even kidding look at this look at the wheels on it it's incredible and the things that you go to the combines in Iowa there's satellite driven today that's legal today is we're already past that hurdle right you can map your property down to the square inch you get these things to come on line at 2:00 a.m. when it's colder outside less stress on the plants you're asleep are you kidding it's awesome it's unbelievable this stuff's happening already today and the same things happening in the mining industry we've got these insane trucks autonomously driven right they're more accurate than humans incredible so really all we're waiting for is is regulatory approval this looks gonna happen really quick I bet 2020 to 2023 it's going to start and if I again if wrong you can call me out I've been wrong many times but I think there's a lot of reasons to believe 2023 is when it starts it'll start in the city centers take a while to get to rural America keep in mind companies like uber and Google are gonna take Wall Street money and buy a hundred thousand cars at a time it's not like we have to wait for individuals to replace their cars with electric autonomous car that's not the way it's gonna work uber will come in and buy a hundred thousand vehicles and just flood the streets right and we're done it's over right the whole thing happens really quick and it's 70 percent of the it's not I mean you know it doesn't go from 1 to 100 overnight but I bet you anything it'll go from 1 to 50 or 60 within 5 years within a you know think like San Francisco or New York or something like that that could go quick so and then you have natural language processing which is another huge piece and I know many of you are working with natural NLP they call it NLP because there's all sorts of applications there Siri was introduced of course in 2011 and we've got Amazon Alexa and that came out in 15 and now of course we've got Google home so just out of curiosity how many of you here today have either Amazon Alexa or Google home in your homes look at that that's incredible yeah that's awesome so where's the biggest financial incentive that's what you always have to you know to see where it's going to propagate first it's always where is the largest financial and center an incentive and the call centers you know I'm sure many of you have huge call centers and there's companies that have like thousands even tens of thousands of people working in call centers what a nightmare to manage that and hire all these people and they're calling in sick and you got a trained em and consistent protocol delivery it's a nightmare imagine how much easier would be to have a computer run that and we're close to it we're getting close to that right but it can also be used for medical billing data entry whether it's written notes by doctors or recorded messages it can be used in in that capacity it's being used to identify people who might be suffering from depression and anxiety even suicide because they're finding patterns in the way people speak they can they can hear a speech pattern right there you can do an automate interview and the way people answer questions I don't know what it is they're seeing but the way people speak can be a very good indicator that they might be suffering from depression incredible right so these are all super disruptive technologies right so let's let's and again I you know I've got basically twenty twenty minutes left so let's take a few minutes and talk about blockchain as well and I know that many of you know a lot about blockchain already which is awesome and I certainly don't want to be repetitive but but let me just obviously block change the technology behind Bitcoin I'm not going to talk about that although Bitcoin is more than doubled so far this year so it's doing well from its low but I want to talk just briefly about structurally how blockchain works and again I know some of you know this but let you stay with me on this because it's a lot of people explain it as if though it's like this really complicated thing and it's not really it's quite simple and in many ways so in the in the simplest analysis just imagine an Excel spreadsheet right but a with transactions listed on it but instead of that spreadsheet being on one computer it's actually on hundreds or thousands of different computers simultaneously they're called nodes right and all those nodes have to agree through a consensus mechanism that they all have the latest most up-to-date version right so it's effectively a distributed ledger but and it makes it really hard to hack because you can't just hack one node and get away with it you have to hack like 51 percent of the nodes at the same time in the same way all successfully in order for the hack to actually be successful but it gets even better than that because every new and here's the essence of blockchain every time a new block is added to the chain of previous blocks right so a new spreadsheet is added to the chain of spreadsheets right effectively the latest block has a summary of what was on the previous block it's called a hash right which means that once something is on the blockchain you can never change it because now you have to you have to hack every subsequent block on 51% of the nodes all of the same time all in the same way all successfully for the thing to actually work and and just if you're curious nobody to date Bitcoin blockchain launched in 2009 no one has hacked the Bitcoin blockchain they've hacked wallets and exchanges that interact with the Bitcoin blockchain but no one has ever successfully and believe me everybody has tried its ton of money because remember they're the just recently the CEO Canadian guy CEO of a company he went to India to do some humanitarian work he ended up passing away over there he got sick and passed away he was the only one who had the key to a wallet that is a hundred and ninety million dollars in it no one can get at it you can see it everyone can see it it's on the public blockchain no one can get at it like that's how crazy is that same thing is true for the DAO hack in 2014 or 2015 I can't remember 50 million bucks it's still sitting there no one's touched it everyone can see it it's crazy so effectively blockchain is just a software architecture it all it is a software architecture records transactions in a new way but it does it in such a way that it automates trust so blockchain automates trust right it does it on its own there's no human involved it's just just a way of recording transactions but the way it's done makes tampering and and and hacks virtually impossible okay so POCs on that one of the early winners with blockchain was in supply chain management and I know you guys are doing some stuff with digital identity but this is an interesting POC evolved 88 bales of cotton right produced in Texas and shipped over to China and all the details of that transaction were uploaded to a private etherion based blockchain using what's called a smart contract right smart contract is just a contract that has if-then statements in it so if you get here if a happens go this way if B happens go that way and then you've got another one if C happens here D and all the permutations may see a decision tree just a piece of code but it's got all the permutations all the possible outcomes of an order and they upload that to the blockchain so no one can change it right and then when when the the the cotton gets all the way to China and the barcodes are scanned that automatically transferred ownership and releases payment it just happens on its own there aren't any people involved right so if you want to understand blocked our smart contracts just imagine wills and and and contracts that effectively execute themselves but this is a huge implication for supply chains because a transaction like that right so you've got the the the company you got the producer of cotton right you've got a trucker to get it to the port you got loader that's a different company it's probably customs clearance at that point now you're on the boat there's probably a dozen companies involved in that one way or another and on the other side you've got an offload you got another customs clearance and you got trucker so this probably 20 companies involved in the transaction one way or another and every one of those companies just like all your companies have order entry people and order trackers and project management people and and of course after the fact you've got audit and compliance people those are all salaries right those are people getting paid to do that work and everyone that's baked into the pricing and everyone's got their hand in the cookie jar so there's an enormous amount of cost that just goes to all these transactions all these people that are involved you can replace virtually all of it maybe not all of it but you you slowly but surely you can't approach that over a period of time so you can have all the individual participants in the transaction they would all be maintaining their own private node on this private blockchain so when an order is uploaded everyone looks at it and says yeah this is what we agreed to upload is now it's on the blockchain so they're all connected to the blockchain and if you pay for this stuff at the end using a cryptocurrency instead of a bank you can eliminate the bank altogether it's really crazy there's huge implications for that right and there's a lot of people who are getting close to deploying when it comes to even like in the car business you can have different people connected through a blockchain or if you're in the real estate business you can have people can I do the blockchain really fascinating you can monitor the transactions as they go through because people are checking in as different milestones are reached so there's a start-up that came up that was doing a blockchain based tracking ledger for diamonds they were you know people are trying to avoid blood diamonds and the like so they're there you're following precious stones from the mine all the way to retail and the whole thing worked so well and people loved it so much that now that same company is doing the same thing to monitor rx medications pharmaceutical medications they're doing it in Europe to track GMO foods right blockchain can be used to keep track of people's degrees and certifications right it's an interesting application of course it can be used to monitor land ownership and title right and recently I know there's a few folks here from Accenture Accenture did a project in conjunction with Microsoft build in the ID 2020 which was a pretty astonishing it's basically a biometric database is similar in a sense to what you guys are building with blockchain right now there's some similarities but the bottom line is these is mostly for Syrian refugees so so they could with with nothing more than a been a smartphone they could prove who they are and where they came from and so that entitles them to certain benefits and you want to make sure you're not giving the same benefit to the same person twice and the whole thing so that's what you know in the use cases and so in insurance you have a huge kyc universe of applications as well so these are again examples of disruptive innovation right things that are just disrupting and then this is and and this is the last thing I'm going to touch on and then I've got just a few slides at the end so just you know you want to get an idea where we're at but because there's two types I mean there's I would say largely there's two types of innovation there's more than that but there's two big buckets and it's incremental innovation and disruptive innovation and incremental innovation that comes like from the center of expertise right so you've got specialists and experts and scientists and people that are calibrating and making things a little bit better every year year after year and this is incredibly powerful like our world we are optimizing our planet and it's a function of incremental innovation I'm not denigrating this at all it's an incredible process that's happening but that's not disruptive innovation right disruptive innovation invalidates existing business models like it comes in with something so new and so different it just pulls the rug out from an entire in soap classic example who developed the digital camera Kodak imagine Kodak the very company that got wiped out by digital cameras they developed it but they looked at in there like oh this thing's a piece of garbage no one's gonna want it it's big and clunky expensive super low resolution they didn't understand the exponential nature of these things okay so they let it go and then it went off and the story getting better and better or better next thing you know it comes back and takes them out crazy right so you have these kind of again classic examples there's tons of them Apple disrupted the music business right Apple disrupted the phone business and then of course Google disrupted the phone business as well so disruptive innovation comes from what I refer to as adjacent markets right it comes from the sides it doesn't come from the center it comes from the sides right what's in adjacent market think about what's your biggest supplier who else do they sell to that's an adjacent market right who's your biggest customer who else do they buy from that's in adjacent market so you have these these industries that are kind of side by side but they're not the same but they kind of overlap in places right and the disruptive innovation comes from those overlaps right they come from the sides that's why people miss them these are smart people some of these executives and you know fortune 500 companies my gosh there's smartest people in the world quite often they got caught off guard because they're focused on this Center right that's it they call it an institutional blindness by the way there's a whole study on this the people after just three weeks in a given organization you start to lose perspective on things outside of your space and that's where the disruption comes from so you again tons of examples right the Facebook messenger it disrupted the SMS business and of course that's what whatsapp did as well and that's what snapchat is doing LinkedIn disrupted the recruiting business right that's an adjacent market right Amazon disrupted the book business with their Kindle e-reader it wasn't the first eReader so Jeff Bezos that guy for Amazon he's brilliant it's incredible health so he said the only thing that is truly disruptive is customer adoption because if people don't adopt it no disruption happened right so there were other ear eaters before the Kindle that they never got broad market adoption he brings out the Kindle they bring out the Kindle broad market adoption all of a sudden the whole book business is different it was an adjacent market right in the burrs disrupting the food delivery business right is a huge thing in China they've got huge and then Amazon's getting into the grocery business right Tesla and Google getting into the car business 15 years ago no one would have believed that and here we are and Facebook wants to disrupt the isp business with these guys these planes they they're they're solar-powered they fly at 70,000 feet they can be up there for months at a time with internet access to the people below unbelievable and Google's trying to do the same thing with the project limb and just last week StarLink it's owned by Tesla sent up 60 satellites they're gonna create a constellation of 12,000 satellites guys if now he pulls that off he's got a lower cost structure than anyone else in Rockets so it's cheaper for him to get him up there he's gonna have internet access across the whole world Elon Musk could be one of the most powerful people on this planet if he pulls that off you think Huawei is tricky you would have the whole thing it's incredible right so so disruptive innovation comes from those overlaps right and you say and it's probably in places where people don't want to look or they they think it's like the crazy sketchy stuff on the side so they don't bother looking so what's the the bleeding edge the cutting edge of nutritional supplements right think about what's the the leading edge of nutritional supplements it's in horse racing right that's where they test the latest sup and makes sense it's less regulated you got a million bucks on the horse he's not quite good enough to win you'll try anything to see if it'll work that's where a lot of that stuff that's where they that's where they tested that's where creatine started some of you probably take I take creatine that's started in horse racing right that's where they test to d3 or the the impact of zinc loading on testosterone that all started in horse racing it's not the only place but it's one of the places where things happen and then this is when it proves that it's got some effect so it starts to migrate in right what's the cutting edge of online marketing guys it's in porn right it's an online gambling right that's where they test the latest wireframes and lead magnets and conversion funnels all that stuff happens in those kind of sketchy parts of the internet that no one wants to talk about but that's where all the Guru's that are selling thousand-dollar info products those guys do all those stuff first trip wires and all that stuff it all comes over there so disruptive innovation comes from the sites and there's one more I want to I want to touch on and you all be done in eight minutes here but but fascinating so this comes from that book innovators dilemma was like I forget the author's name what's the author of that book innovators dilemma nope okay doesn't matter in a disruptive innovation tends to cater to the least profitable market segment first right so think about your your business what's your least we're talking about the customers you hate the most oh there's such a pain in the neck they have no money they call twice as often as anybody else you know the ones right every industry has them you almost be happy if they'd go away entirely right that and what do they want they want simpler less expensive solution right that's disruptive innovation in your industry simple or less what do people really do and this is true in every industry they cater to the most profitable industry segments market segment which is at the top they're selling premium products which is good there's money there there's good profits there you're selling premium products service packages all sorts of stuff that's high margin stuff you need that money right because innovation costs okay innovation can be really summarized in one concept two words budgeting failure you had a budget failure where you go to Jeff Bezos again if you know it's gonna work it's not an experiment I mean this stuff's brilliant right if you want to innovate it implies you're trying something when you don't know if it's gonna work or not which means it might fail which means you need a budget for it where do you get that money from you get that money from selling premium products right that's where the money comes from so you have this pervasive upward push it happens in every industry as everyone's trying to sell these premium high-end products and who gets left behind raise the people at the bottom right the least profitable market segment that's where disruptive innovation comes from so I do these strategy sessions all the time with these with with with with you know and with company people and I always tell them to look up look down look side-to-side right look up that's where the profit is right that's where the the premium you need gross profit you need gross margin right that stuff is super important you need that money if you're gonna innovate right but then at the same time look down that's where disruptive innovation comes from it could do a meeting once a month on a Friday afternoon when things are slow and just just have a brainstorming session what do these people want who's playing who's trying to capture that business right and then finally it looks side to side to those adjacent markets they come back to where we started from who else's lunch can we eat right that's those are new revenue opportunities right stay on offense we got to stay on offense what can you do with data analytics right what can you do with machine learning and how what can you do with with with blockchain my message for you today is excuse me it's just to think way bigger about what's possible and what you could be doing right and what your competitors could be doing and when you think bigger like there's a few things you can kind of take for granted the first one is that you end up inspiring everyone around you you have a big what was the guy in the panel said to say the bricklayer and the second guy was like yeah I'm picturing the the castle on the roof it's inspiring to him right that's the whole point you have a big goal like that you have a some big goal that you're doing here we're gonna revolutionize our industry you want to be a part of it people like yeah absolutely ray your employees get your customers get inspired even your competitors get inspired when you're doing something big and if you do something that no one else has done you've got knowing no competition because no one else has done that yet so there's a whole there's a ton of people obviously that are known for kind of thinking this way I'm a big you know I love quotes I literally buy books of quotes and this guy's got like thousands of great quotes so here's my favorite richard branson quote fastest way to become a millionaire is to start out as a billionaire and then start an airline [Laughter] that's genius anyway so he started virgin galactic like how much bigger are you gonna get the man right any competition not much then you got Elon Musk I mean this guy's like a hero to me I mean I I know he has his haters and there's 200 people I'm sure some of you don't like him and that's cool I love him I mean to me he's like I it's hard to believe that he exists I mean I just it's like he spends his entire life doing impossible things but anyway I mean he so he started Tesla started SpaceX by the way how many of you honestly how many of you watched is the Falcon Heavy has launched twice how many of you have watched either live or on YouTube even one of those two launches I love you guys and for the rest of you if you're if you're into this even that much like if you're a nerd that much go to youtube and search for Falcon Heavy launch it's a 12-minute video right it's because they're reusing the hardware right they're sending the rocket up and then they take these boosters back down and land them simultaneously like pencils on the landing pad right where they took off from 11 minutes earlier it's so great I'm telling you I'm a nerd it's obvious okay but it's true this stuff puts shivers down my spine how do you do that how do you because here's the thing the the okay cost of sending the Falcon nine to the ISS the International Space Station it's about sixty million bucks sixty million bucks how much is the fuel top line is 60 million how much is the fuel I guess 10 million that was my guess 10 million bucks the right answer is $200,000 one 300th of the total cost 1/3 which means if you can reuse the hardware you can bring the cost function down a hundredfold do you know how crazy that is it's insane and and now so the cost function I mean NASA is paying a hundred and thirty three million dollars per launch by contract if they got if they're able to reuse the the fuselage and the the booster their cost functions like 20 million bucks that means they make a hundred and ten million profit because they've got superior technology they just priced it underneath the competition that's all they needed to do to get the order didn't matter that the cost was down here that means they make a fortune what does that money go to it goes to StarLink it goes to the bfr it goes to starship it goes to all this crazy innovation they're funding it does this make sense this is where the whole game is about right you got to get your premium products so you can get your hands on some gross profit and then you got a budget failure and innovate like crazy and it's just gonna get faster and faster so the first time they pulled this off was right just before Christmas on in 2015 and when they pulled it off their employees partied like it was 1999 can you believe it 27 27 years old average age that makes them Millennials Millennials get a bad rap don't they lazy apathetic entitled it's not true those Millennials will work harder than anyone you've ever hired if you give them something inspiring to work towards they're not Millennials they're just young people right we won't we all wanted to do that when we were in our 20s and 30s right anyone who's under 40 is a millennial they're not kids anymore they're half the workforce right they just they want to do something cool right they want to change the world or populate another one so crazy so all these stories right we've all heard these stories before at the end of the day is it my presentation isn't really about technology per se it's not about artificial intelligence or blockchain it's about leadership who's willing to see the market structure for what it is right you need gross profits so you can innovate and innovate life run towards it right stay on the right side of that leverage equation at all cost because otherwise you're swimming upstream right and just think way way bigger about what's possible so anyway I you know it's a crazy day for me I flew in this morning I don't even have a hotel and flying out tonight I'm back home on a Newport Beach I just moved to Newport Beach last August so beautiful but I'm delighted to be here and I'm gonna be here until the cars picking me up at ten after three my favorite part of my job is is talking to people after and sometimes people correct me sometimes people have stories so they have questions or anything like that I would love to hear it I'll be hanging out and I really appreciate your attention for my session thank you [Applause]

Related Posts

1 thought on “Artificial Intelligence in Healthcare Keynote Speech

Leave a Reply

Your email address will not be published. Required fields are marked *